Certain kitchen-middens in Jamaica

51
N4
NMAI fttF

EYE M US

CERTAIN KITCHEN - MIDDENS IN

JAMAICA

■BY THEODOOR DE BOOY:

ii

MVSEVMoftheAMERJCAN INDIAN.'

^;,ii,i:i.jiiiii„iiiiin. :li.:i..m ^lii|Mi|iiill,'iliiliiil,l,illlliLlilllil|)jtfll%lWl''lii,

<I

■f^r-

%^^n :

lois

CONTRIBUTIONS FROM THE HEYE MUSEUM
NUMBER 3

CERTAIN KITCHEN-MIDDENS IN
JAMAICA

By THEODOOR DE BOOY

Reprinted from the American Anthropologist (n.s.), Vol. XV, No. 3,
July-September, 191 3

Lancaster Pa., U. S. A.
The New Era Printing Company
1913

CM

.^THSO/V^

JUL 2 0 2000

[Reprinted from the American Anthropologist, Vol. 15, No. 3, July-Sept., 1913.]

CERTAIN KITCHEN-MIDDENS IN JAMAICA
By THEODOOR DE BOOY

DURING a sojourn on the island of Jamaica in the months
of January, February, and March, 1913, in the interest of
the Heye Museum of New York City, the author was
enabled to excavate some of the typical aboriginal kitchen-middens
found in various parts of the island. These middens shed much
light on the mode of life of the pre-Columbian inhabitants, and,
above all, allow a practically exact determination of their different
foodstuffs. At the same time, such ceramic fragments and other
artifacts as one can find — and they occur in great quantities —
are most valuable for comparative studies of the prehistoric cultures
that existed in the West Indies.

The modern equivalent of the kitchen-midden is the ash-heap:
where, nowadays, we cast our broken kitchen utensils, bottles, and
empty cans, the natives threw their broken pots and cooking-
slabs, their shells, and such stone artifacts as happened to break in
the making or during use. It can therefore readily be seen what
important conclusions can be drawn from the varied specimens
found in a midden. The author considered himself especially
fortunate in being able thoroughly to examine and excavate a
number of these refuse heaps.

The writer wishes to express his sincere thanks in behalf of the
Heye Museum and himself to the Reverend J. P. Hall, of Brown's
Town, Jamaica, for facilities given him in this work, for generous
permission to excavate ,and for the material from the middens and
from other sources presented to the Heye Museum.

In Dr J. F. Duerden's work on the archeology of Jamaica1
the following paragraph can be found on page 19:

1 Aboriginal Indian Remains in Jamaica, Journal of the Institute of Jamaica, vol.
11, no. 4, Kingston, 1897.

AM. ANTH., N. S., 15 — 29 4~5

4^6 AMERICAN AXTHROPOLOCIST [n. s., 15, 1913

Retreat. — These deposits are on the property Retreat, situated between
Brown's Town and Stewart Town, in St Ann, about four miles from the former.
The land is now owned by Mr Roper, but was formerly in the possession of Mr
Moulton Barrett. Miss Moulton Barrett made a number of investigations at
the place and lent to the Anthropological Exhibition a collection of pottery frag-
ments obtained. . . . The hill or ridge upon which the kitchen-middens are found
is about 1,200 feet high, and six miles from the sea. . . . The elevation in question
was very significantly known by the former owners as " Cacique's Ridge," and
is also known as Little Xigger-ground Hill, while a higher one near has the title
of Big Xigger-ground Hill. These latter names recall the fact that in slavery
days the particular spots were used by the Negroes as provision grounds. . . .
Excavations were made at numerous spots, and, in all, scattered amongst the
upper dark loose earth and fragments of limestone, were broken pieces of pottery,
quantities of land shells, a few specimens of marine shells, and the bones of the
Indian coney and of various fish. This foreign material extended in several
places to a depth of two feet. In one a deposit of partially indurated bluish-gray
ash and charcoal gave evidence of the use of fire by the builders of the refuse-
heaps.

At the time of the author's excavations the Retreat property
was owned by Mr Hall, and since the investigations of Dr Duerden
in 1896 no excavations had been conducted on Little Xigger-ground
hill. Figure in shows the top of the hill. The actual crest of the
hill is almost level, and the middens can be readily seen on the
northern, eastern, and southern slopes, around the crest. The
author determined sixteen middens with accuracy by digging
small test-holes in the soil wherever there was a hummock. There
may be several more middens on the hill, which in the course of
time have altered in shape or the original summit of which has
been removed by the excavations of Dr Duerden and previously
by those of Miss Moulton Barrett.

The western part of the top of Little Xigger-ground hill is a
rocky woodland. There was no proof that it had been used as
part of the village site, nor could evidence of burials be discovered.

Little Xigger-ground hill (fig. 112) is situated about six miles
from the sea. It does not afford a view of the sea, as it rises be-
tween the surrounding hills; but sentinels could readily perceive
the approach of marauding Caribs and give ample warning to
enable escape to the neighboring hills in the interior in case of a raid.
An old road, in reality little more than a footpath, still extends
from the hill to the sea, and is reported to have been made by
the Spaniards. It is quite within the bounds of probability that

DE BOOYj

CERTAIN KITCHEN-MIDDENS IN JAMAICA

427

this road was an aboriginal path, and that the Spaniards after-
ward used and enlarged it for their own purposes. The plain due
north of Little Nigger-ground hill is fertile and suggests an ideal
place for the cultivation of cassava.

\ »■'»■»■! reef:

Q6^ &

Fig. hi. — "Retreat" village and middens.

Excavations in the Middens
Excavations were commenced in the midden marked 1 in figures
in and 113. This midden is on the northern slope of the hill, and,
as can be seen on the plan, is somewhat isolated from the other
middens, there being a very small midden in front of it, but other-
wise no other mounds within 160 feet. Midden 1 is 42 feet long
and 30 feet wide; the crest has a height of 4 feet above the slope.

428

AMERICAN ANTHROPOLOGIST

[n. s., 15, 1913

Excavation in this midden produced the same kind of material as
afterward was found in midden 4, of which a detailed account
follows. As three trenches were dug in midden 4 and only one in

Sea Co&st

Fig. 112. — "Little Negro Ground " hill, with neighboring hills. (Not to scale.) \

the first midden, it is preferable to give the detailed account of the
excavations in the larger refuse heap. It is, however, of interest to
note that such pottery fragments as were found in midden 1 were

Fig. 113.

-Side elevation of "Retreat" village site, looking south, showing mid-
dens on slope. (Not to scale.)

far more brittle and were smaller than those from the other accumu-
lations. This was due probably to the better drainage of the
southern slope of the hill.

Midden 2 is on the southern slope of the hill and did not produce
much material. This midden has a more pointed crest than any of

AMERICAN ANTHROPOIOGIST

S., VOL. 15, PL. XXXI

uM >t-

n. LONG TRENCH CUT THROUGH MIDDEN 4

**c

/'■ CUT IN MIDDEN 4, SHOWING SHELL DEPOSITS

de booy] CERTAIN KITCHEN-MIDDENS IN JAMAICA 429

the others and could not have afforded space for more than a single
hut.

Midden 3 is directly south of midden 2, but about twenty feet
lower down on the slope of the hill. This midden also was not very
productive, but, on the other hand, the sherds were in far better
condition than those from the other sites.

It was in midden 4 (figs, in, 114, 115) that the most extensive
excavations were made. The latter part of these excavations was
not made with the purpose of obtaining more archeological material,
but with the idea of procuring as accurate information as possible
regarding the location of the huts and the exact limits of the shell
and ash deposits.

O Sod and c///vv/a/ deposit.

m Shelf deposit.

fli Ash ef epos it.

E2) Mar/ and /arger atones

of tea* //nes show probab/e /oca //>/

Fig. 114. — Cross-section of midden 4. (Not to scale.)

As in all previous excavations on Little Nigger-ground hill, an
enormous number of land-snail shells were unearthed; indeed, from
test-holes dug in midden 4, the author computes that this midden
alone contained more than 300 cubic feet of shells.

Three trenches were dug in this midden, besides a large number
of test-holes. The first trench (a, fig. 115) was 14 feet long by 10 feet
wide, with varying depth down to marl. This trench extended in a
south to north direction. At the most northerly point the marl
came immediately under the diluvial deposit and the sod, and no
shells, ashes, or artifacts were found. The second trench (b) was
38 feet long by 10 feet wide and followed a southeast-northwest
direction, with a depth of four feet at the southeastern end, down

43°

AM ERIC AX AXTHROPOLOGIST

[N. S.

o. 1913

to marl, and a depth of less than a foot on the crest. This trench
was purposely continued beyond the crest, with a view of verifying
the test-holes previously dug. As will be shown later, this long
trench gave important testimony as a basis for a hypothesis con-
cerning the exact location of the aboriginal abode. The third
trench (c, fig. 115) was 10 feet long and 10 feet wide.

At the point marked b, the author started the second trench
by having the sod and diluvial deposit removed. This was a
foot thick at the base of the midden, where the trench was begun.
Proceeding toward the crest, after a distance of a foot, a layer of
shells was found under the diluvial deposit, which layer gradually

A - first TrencA.

f 3- Second Tre/yc/).

~T C ~ T/?/rd TrencA.

Proiab/e /occrS/oo 0/ //</fa.
-^Arrows t'ne//cafe cftrecf/'on of sAc/Z-t/e/Doj/b
Ash c/eposHs around .Southern s/o/?e.

Fig. 115. — Top view of midden 4. (Xot to scale.)

increased from a thickness of six inches at the very base ■ to a
thickness of a foot about two feet inward, and then tapered grad-
ually to a single shell or two at the summit of the mound. Fol-
lowing the trench down the opposite slope, no shells were found.
Mingled with the shells on the southern slope of the midden were
the artifacts, such as worked stones and pottery fragments, many
of the latter in excellent condition. The ash deposit commenced
four feet above the foot of the slope and continued to the summit.
This deposit was from 8 to 12 inches in thickness, and in it were
found sherds, bones, worked stones, etc. Xo ashes were found on
the northern slope of the midden. Test-holes showed that the ash
deposit extended round the southern slope, covering in all about
half the circumference of the midden, while the shell deposit covered
about 270 degrees of the circumference. On the crest of the midden

AMERICAN ANTHROPOLOGIST

N. 8., VOL. 15, PL. XXXI

a, i. FRAGMENT OF BOAT-SHAPED VESSEL. (THREE-FOURTHS)

FRAGMENT OF CLAY GRIDDLE. (ONE-HALF)

debooy] CERTAIN KITCHEN-MIDDENS IN JAMAICA 43 1

two ash-pockets were found, about two feet in diameter and 12
inches deep, surrounded by stones about the size of a man's head.
The presence of these ash-pockets is problematical, as the aborigines
certainly did not need a fire to keep their abode warm in a tropical
climate, and the semicircle E — D (fig. 115) around the hummock
supports the theory that cooking was done on the slope of the
midden rather than on the crest. It is not from the two ash-
pockets, but from the direction of the shell and ash deposits, that
the author reaches the conclusion that the aboriginal abode was
situated on top of the hummock and has formed the theory that
the inhabitants threw the refuse and the discarded artifacts in three
quadrants of the circle: in other words, in all directions save one,
which afforded them at least one clean, clear space in front of the
hut.

That the native made use of a natural rise of the ground on
which to build his hut can plainly be seen in plate xxxi, a, which
illustrates the long trench (b) cut through midden 4. It can here
be seen that the contour of the midden is practically the same as
that of the surface of the ground, as the trench was cut down to
marl and all loose earth, shells, and ashes had been taken out before
the photograph was made.

A good idea of the shell deposits can be had from figure b of the
same plate, in which a section of trench A in midden 4 is shown. In
all the middens on Little Nigger-ground hill at least 98 per cent, of
the shells found were those of land-snails, the remainder being sea-
shells. The author gathered specimens of the different species
and a later examination identified them as follows :

Pleurodonte acuta
Pleurodonte acuta ingens
Pleurodonte sinuata
Sagda alligans
Area noae
Area dehayesi
Livona pica
Fasciolaria distans
Avicula
Strombus gigas

Various small animal bones were found in the middens, but the

43 2 AMERICAN ANTHROPOLOGIST [n. s., 15, 1913

identification of these has not been possible. It is certain, how-
ever, that some of them belonged to the utia (Capromys). There
also were many fish-bones, and these for the greater part have been
identified as belonging to the rock-fish (Mycteroperca), which at the
present time is regarded as one of the most edible fishes in the
Caribbean.

Archeological Specimens from the Middens
The predominating type of pottery vessel in these middens, as
well as in other middens of Jamaica investigated by the author,
is a boat-shaped type, and while naturally no entire vessel was
found, enough remains of several of them to show their form quite
distinctly. In plate xxxii, a, &, is shown part of a vessel of this type,
made of a dark-red clay, partly blackened either in the firing or by
usage. The wall of the vessel is not more than one-eighth of an
inch thick, and with the exception of the handle, which shows a
few simple incised lines, it is undecorated. The rim is incurving
and has a banded edge.

Many of these boat-shaped vessels have handles that are more
elaborately decorated, and one occasionally finds handles that show
a conventionally modeled parrot's head. Human heads also are
depicted. Plate xxxiii, I , shows some of the handles of these types.
Fragments of decorated rims also were found in abundance, the
decoration consisting chiefly of straight-line incisions, while serrated
lugs under the rim are not uncommon.

The difference between Jamaican pottery and that from the
neighboring islands impressed the author as being quite consider-
able ; this is especially noticeable in the type of handles and in the
fact that in the rectilinear decorations on sherds from most of the
other islands an indentation made with the same tool that made the
incised lines is found, which does not occur on any of the many
hundreds of sherds collected by the author in Jamaica. This feature
is described by Dr J. Walter Fewkes1 as follows:

A marked feature in rectilinear decoration is the indentation of the extremity
of each line. The potter commonly terminated a line with a shallow pit that was

xThe Aborigines of Porto-Rico, Twenty-fifth Annual Report, Bureau of American
Ethnology, Washington, 1907, p. 180.

AMERICAN ANTHROPOLOGIST

N. S., VOL. 15, PL. XXXIII

1. HANDLES OF VARIOUS TYPES. (TWO-THIRDS)

2. SHERDS WITH INCISED DECORATION. (ONE-HALF)

3. TYPICAL JAMAICAN HANDLES. (TWO~THIRDS)

debooy] CERTAIN KITCHEN-MIDDENS IN JAMAICA 433

apparently made with the same instrument as the line itself: or it was sometimes
slightly separated from the end of the line. So constant, almost universal, is
this feature that it may be looked on as characteristic of pottery from Porto Rico
and Santo Domingo.

This feature was also noted by the author in the sherds found
on the Caicos islands and the Bahamas, which, on comparison, can
hardly be distinguished from Santo Domingo sherds. Dr Fewkes
also says (page 188) in the work referred to:

Illustrations b, c, and d are representations of fragments of pottery from Nipe
Bay, Cuba, and show the striking resemblance between the ceramics of Porto
Rico and those of the largest of the Antilles.

It would therefore appear that the Jamaica pottery should not
be included in the culture-type of that of Porto Rico, Hayti, Cuba,
and the Bahamas, but that it belongs in a class by itself. It is
rare to find impressed decorations on Jamaican pottery, and in the
few instances in which it is found, the ornamentation is crude. On
the other hand, as can be seen in plate xxxiii, 2, some of the
incised decorations are fairly elaborate.

The middens of Little Nigger-ground hill also produced a type
of handle that is not found in the other West Indian islands. Dr
Duerden figures one of this type in his monograph above referred to,
and it seems to be found in middens in various parts of Jamaica,
with only slight variations in size and decoration. Plate xxxiii, 3,
illustrates a few handles of this type. Whereas all the pottery
previously described has a uniform thickness of from one-eighth
to three-sixteenths of an inch, the thickness of the vessel to which
this type of handle belongs is at least a quarter of an inch, and
increases to not less than half an inch at the back of the perforation.
Raised, serrated ornamentation is shown under the perforations in
some cases (pi. xxxiii, 3, a) ; in other specimens two small lugs were
added by the potter (c) ; again, a kind of scrollwork (b) is employed,
as if the potter endeavored in her crude way to represent the ten-
tacles of an octopus; while in d and e, two fragments belonging to
the same vessel, an attempt may have been made to indicate a
mouth, belonging to a face, of which the ridge between the perfora-
tions would represent the nose and the perforations the eyes.

434 AMERICAN ANTHROPOLOGIST [s. s., 15, 1913

Besides earthenware vessels, the middens also produced many frag-
ments of clay griddles (pi. xxxn, c). These are about an inch thick
at the base and taper to a thickness of half an inch in the middle.
They had a diameter of about 14 inches, but one finds only com-
paratively small fragments of these objects, which more than likely
is due to the fact that the action of fire made them more brittle
than the other ceramic objects.

The author found the usual worked stones in the middens,
consisting of celts of the petaloid type, hammerstones, and rubbing
and smoothing stones. A single specimen of shell celt was found,
fashioned from the lip of a conch. ,

Heye Museum

New Vork City